Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A system for document processing comprising: means for providing a scanned image of a document to be processed, the scanned image obtained by a processor performing optical character recognition (OCR) on the document; means for decomposing, using a computer, the image of the document into at least one sub-image, the sub-image defining a character-based recognition element; means for providing the at least one sub-image from the computer over a network to a plurality of data entry computers when a confidence rating related to an outcome from performing OCR on the sub-image is below a predefined threshold, the sub-image being displayed on one or more data entry computers so as to allow one or more user-based inputs, in response to observation of the displayed sub-image, to provide one or more user-based data entry values respectively from at least one of the data entry computers, the one or more user-based data entry values being communicated over the network to a receiving computer; means for determining, by the receiving computer, if the one or more user-based data entry values provide a predetermined level of confidence; and means for computing, by the receiving computer, a verified result when the predetermined level of confidence is provided.
A document processing system automates data entry from scanned documents. It uses OCR to create a digital image of the document. This image is divided into smaller sub-images, each representing a character or data field. If the OCR confidence score for a sub-image is low, the sub-image is sent over a network to multiple data entry clerks. These clerks view the sub-image and manually enter the correct value. The system receives these entered values and determines if they meet a confidence threshold (e.g., multiple clerks enter the same value). If the threshold is met, the system considers the result verified.
2. The system of claim 1 wherein the verified result identifies the character-based recognition element when the confidence rating related to the outcome from performing OCR on the sub-image indicates the sub-image was not successfully processed.
In the document processing system, if the OCR process fails to accurately recognize a character-based element in a sub-image, the verified result from the data entry clerks identifies and replaces the unrecognized element. This ensures that even if the initial OCR process is unsuccessful, the correct character is still captured through manual verification and data entry.
3. The system of claim 1 wherein the verified result confirms the character-based recognition element when the confidence rating related to the outcome from performing OCR on the sub-image is below the predefined threshold.
In the document processing system, if the OCR process produces a low confidence score for a character-based element, indicating a high likelihood of error, the verified result from the data entry clerks confirms the correct element. The verification step, using human input, boosts confidence in the accuracy of the final processed document.
4. The system of claim 1 further comprising: means for communicating the one or more user-based data entry values over the network to the receiving computer.
The document processing system includes a means to transmit the user-entered data values from the data entry computers to the central receiving computer over the network. This communication channel allows the system to collect and aggregate the data entered by multiple users for verification and further processing.
5. The system of claim 1 wherein the network comprises the Internet.
The network used for communication between the data entry computers and the receiving computer in the document processing system is the Internet. This allows for distributed data entry from anywhere with an internet connection.
6. The system of claim 1 wherein the computer that performs the decomposing includes the receiving computer that performs the computing.
In the document processing system, the computer responsible for dividing the document image into sub-images is the same computer that receives the data entry values and computes the verified result. This consolidates processing onto a single machine.
7. The system of claim 1 further comprising: means for collating the verified result into a character-based electronic document corresponding to the document image.
The document processing system compiles the verified results from data entry clerks into a complete, searchable electronic document that corresponds to the original scanned image. This ensures the creation of an accurate and complete digital representation of the physical document.
8. The system of claim 1 further comprising: means for performing OCR on the sub-image to generate an OCR value; and means for determining whether the OCR value matches the verified result.
The document processing system uses OCR to generate an initial value for each sub-image. The system then compares this OCR value against the verified result provided by the data entry clerks. This comparison allows for quality control and identification of instances where OCR failed and manual correction was necessary.
9. The system of claim 1 wherein means for providing the at least one sub-image comprises providing the at least one sub-image to at least one of the one or more data entry computers, wherein a user at the at least one of the one or more data entry computers is known to be available for viewing the sub-image.
The document processing system sends sub-images to data entry clerks who are known to be available and ready to review and process the images. This optimizes workflow by ensuring that tasks are assigned to active users, reducing turnaround time.
10. The system of claim 1 wherein means for providing the at least one sub-image comprises providing the sub-image to a user at one or more of the one or more data entry computers in connection with a work order.
The document processing system provides sub-images to data entry clerks in connection with a specific work order. This allows for task management and tracking of progress for individual documents or batches of documents being processed.
11. The system of claim 10 further comprising: means for determining that the user at the one or more data entry computers is available if a profile associated with the user matches a profile associated with the work order.
The document processing system assigns work orders to users based on matching profiles. The system checks if the user's profile aligns with the work order's requirements before assigning the task, ensuring that appropriate personnel handle specific document types or data entry needs.
12. The system of claim 1 wherein the means for determining includes determining whether a predetermined number of the user-based data values of a plurality of the user-based data values match so as to provide the predetermined level of confidence.
The document processing system determines whether a sufficient number of data entry clerks agree on a value for a sub-image to reach a predetermined level of confidence. For example, if three out of five clerks enter the same value, the system considers the result verified.
13. The system of claim 8 wherein means for determining includes determining whether a single one of the user-based data values matches the OCR value and if the single one of the user-based data values matches, returning the verified result.
The document processing system compares the OCR value with the data entry values received from clerks. If a single clerk's entry matches the OCR value, the system immediately accepts that entry as the verified result, streamlining the verification process when human input confirms the initial OCR output.
14. A system for document processing comprising: an OCR processor that performs OCR on a document so as to provide a scanned image of the document; a decomposing computer that decomposes the image of the document into at least one sub-image, the sub-image defining a character-based recognition element; one or more user-based data entry values that are provided respectively from at least one of a plurality of data entry computers a receiving computer that receives one or more user-based data entry values over a network from a data entry computer that views the sub-image, the sub-image being provided from the decomposing computer over the network to a plurality of data entry computers when a confidence rating related to an outcome from performing OCR on the sub-image is below a predefined threshold, the sub-image being displayed on one or more data entry computers so as to allow one or more user-based inputs, in response to observation of the displayed sub-image, to provide the one or more user-based data entry values that are communicated over the network to the receiving computer; determines if the one or more user-based data entry values provide a predetermined level of confidence; and computes a verified result when the predetermined level of confidence is provided.
A document processing system uses OCR to create a digital image of a document. A computer then divides this image into sub-images representing characters or data fields. If OCR confidence is low, these sub-images are sent to data entry clerks for manual input. The system receives these inputs and determines if they meet a confidence threshold (e.g., majority agreement). If the threshold is met, the system considers the result verified and saves the verified data.
15. The system as set forth in claim 14 wherein the decomposing computer includes the receiving computer and the processor.
In the document processing system, a single computer performs the tasks of decomposing the document image into sub-images, receiving user data entries, and performing the initial OCR processing. This consolidated setup streamlines the workflow.
16. The system as set forth in claim 14 wherein the decomposing computer is a separate computing device from the receiving computer.
In the document processing system, the computer that decomposes the document image into sub-images is a separate machine from the computer that receives the data entry values and computes the verified result. This distributed setup allows for dedicated resources for each task.
17. The system as set forth in claim 14 wherein the network comprises the Internet.
The network used for communication between the data entry computers and the receiving computer in the document processing system is the Internet. This allows for distributed data entry from anywhere with an internet connection.
18. The system as set forth in claim 14 further comprising means for collating the verified result into a character-based electronic document corresponding to the document image.
The document processing system compiles the verified results from data entry clerks into a complete, searchable electronic document that corresponds to the original scanned image. This ensures the creation of an accurate and complete digital representation of the physical document.
19. A system for processing scanned images, the system comprising: an OCR processor that performs OCR on a document so as to provide a scanned image of the document; a decomposing computer that decomposes the image of the document into at least one sub-image, the sub-image defining a character-based recognition element; means for providing the at least one sub-image from the computer over a network to a plurality of data entry computers when a confidence rating related to an outcome from performing OCR on the sub-image is below a predefined threshold, the sub-image being displayed on one or more data entry computers so as to allow one or more user-based inputs, in response to observation of the displayed sub-image, to provide one or more user-based data entry values respectively from at least one of the data entry computers, the one or more user-based data entry values being communicated over the network to a receiving computer; means for determining, by the receiving computer, if the one or more user-based data entry values provide a predetermined level of confidence; and means for computing, by the receiving computer, a verified result when the predetermined level of confidence is provided.
A system processes scanned images. It uses OCR to create an image of the document and then divides the image into sub-images. If the OCR confidence is low for a sub-image, it is sent over a network to multiple data entry clerks. These clerks view the sub-image and manually enter the correct value. The system receives these entered values and determines if they meet a confidence threshold (e.g., multiple clerks enter the same value). If the threshold is met, the system considers the result verified.
Unknown
November 18, 2014
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.